Social Impact Bonds a Report to the Alaska State Legislature

Social Impact Bonds A Report to the Alaska State Legislature

December 15, 2016

A review of Social Impact Bonds and their potential to reduce recidivism rates in Alaska.

The Alaska Criminal Justice Commission

http://www.ajc.state.ak.us/alaska-criminal-justice-commission

ACJC Social Impact Bonds Report

Table of Contents

Executive Summary ...... 1 Introduction ...... 3 Structure and Mechanics of Social Impact Bonds ...... 3 Examples of Social Impact Bonds in the Field ...... 5 Peterborough Prison Project ...... 5 The New York City Rikers Island Project ...... 6 Utah High Quality Preschool Project ...... 6 Massachusetts Chronic Homelessness Pay for Success Initiative ...... 7 Massachusetts Juvenile Justice Pay for Success Initiative ...... 7 Anchorage/Mat-Su Permanent Supportive Housing Demonstration ...... 8 Preconditions of Social Impact Bonds ...... 9 Target prevention ...... 9 Reduce budgetary pressure ...... 9 Grow recipient population ...... 10 Adhere to evidence base ...... 10 Avoid creating a dedicated fund ...... 10 Legislative Research Topics ...... 11 Research Topic 1: Identification and evaluation of grant programs, contracts, and services of the Department of Corrections and the Department of Health and Social Services that may be suitable for social impact financing ...... 11 Research Topic 2: The possibility of private sector investors providing social impact financing15 Research Topic 3: Programs operated by nonprofit corporations that could be funded through a social impact financing mechanism ...... 20 Research Topic 4: Independent evaluators that could determine whether performance targets for a nonprofit corporation funded by social impact financing are met at the end of an agreed-upon time frame ...... 22 Research Topic 5: Whether federal funding is available for independent evaluators participating in social impact funding ...... 25 Conclusion and Recommendations ...... 25 Works Cited ...... 27

ACJC Social Impact Bonds Report

Executive Summary

As part of SB91 (Omnibus Crim Law and Procedure; Corrections of 2016), the Alaska Legislature requested a report regarding the The Alaska Criminal potential of using social impact bonds to reduce recidivism rates Justice Commission in Alaska. Social impact bonds are a form of performance-based contracting, in which private or philanthropic organizations fund The Alaska State Legislature created the Alaska Criminal Justice the delivery of a program on the condition that, if certain Commission in 2014. programmatic success measures are met, expenses will be The Commission consists of 13 reimbursed by the state. To evaluate their potential in Alaska, five members: research topics were proposed by the Legislature; these follow. o Gregory P. Razo, 1. Identification and evaluation of grant programs, Chair, representing the Alaska Native Community contracts, and services of the Department of Corrections o Alexander O. Bryner, and the Department of Health and Social Services that designee of the Chief Justice may be suitable for social impact financing: o John B. Coghill, Senate, Non-Voting - Fifteen programs demonstrated a positive impact o Wes Keller, House, on recidivism and were sufficiently evidenced Non-Voting (until Jan. 2017) Jahna Lindemuth, based. o Attorney General o Jeff L. Jessee, 2. The possibility of private sector investors providing social Alaska Mental Health Trust impact financing: Authority o Walt Monegan, - There are private and philanthropic organizations Department of Public Safety operating in Alaska whose mission overlaps with Commissioner Stephanie Rhoades, programs suitable for social impact bond o District Court Judge financing. Additionally, there are several o Kristie L. Sell, organizations who, it has been announced, have Municipal Law Enforcement o Brenda Stanfill, expressed interest in a program financed by a Victims’ Rights Advocate social impact bond in Alaska; these include, o Quinlan G. Steiner, Public Defender Alaska Mental Health Trust Authority, Mat-Su Health o Trevor N. Stephens, Foundation, and the Rasmuson Foundation. Superior Court Judge o Dean Williams, Department of Corrections 3. Programs operated by nonprofit corporations that could Commissioner be funded through a social impact financing mechanism:

ACJC Social Impact Bonds Report

- Four programs demonstrated a positive impact on recidivism and were sufficiently evidenced based. These included: Reentry Services; Housing Assistance; Cognitive Behavioral Therapy; and Employment and Job Assistance.

4. Independent evaluators that could determine whether performance targets for a nonprofit corporation funded by social impact financing are met at the end of an agreed- on time frame: - Several organizations and institutions provide program evaluation services in Alaska.

5. Whether federal funding is available for independent evaluators participating in social impact funding: - Federal funding is currently unavailable for evaluation services.

While several projects utilizing a social impact bond have ended, as a financing structure, social impact bonds are still largely untested. Given this, the Commission recommends to the legislature that it not actively solicit social impact bonds, as that would require expending state resources on a mechanism that is currently neither well-documented nor evidence-based. However, if a private entity were to submit a social impact bond proposal to the state with the majority of legal and administrative technical components resolved, the Commission recommends accepting such a proposal following appropriate vetting.

ACJC Social Impact Bonds Report

Introduction

In SB91 (Omnibus Crim Law and Procedure; Corrections of 2016), the Alaska Legislature directed the Alaska Criminal Justice Commission (Commission) to produce a report regarding “the potential of using social impact bonds to reduce recidivism rates.”1 The legislature further directed the Commission to address the following topics:

(1) identification and evaluation of grant programs, contracts, and services of the Department of Corrections and the Department of Health and Social Services that may be suitable for social impact financing; (2) the possibility of private sector investors providing social impact financing; (3) programs operated by nonprofit corporations that could be funded through a social impact financing mechanism; (4) independent evaluators that could determine whether performance targets for a nonprofit corporation funded by social impact financing are met at the end of an agreed-on time frame; and (5) whether federal funding is available for independent evaluators participating in social impact funding.2

This report squares the legislative request with what is known about social impact bonds. As a new financial tool, there is a dearth of information on the subject despite burgeoning interest. This report focuses on what can be ascribed to the evidence base. To this end, the report begins with an overview of what social impact bonds are, how they operate, and examples of their use in the field, including a planned program in Alaska. Preconditions of social impact bonds follow. Finally, the research topics that the Legislature identified are discussed.

Structure and Mechanics of Social Impact Bonds

Vulnerable populations in Alaska often face complex social, economic and health problems. When well implemented, programs that target these groups may reduce the need for subsequent remediation, for example, reincarceration or special education. To date, it has most often been government agencies and nonprofit institutions, often in collaboration, who have identified needs and provided services to vulnerable members of the community. Government agencies and

1 Omnibus Crim Law and Procedure; Corrections, AK SB91, 36 SLA 16 (2016), Retrieved from http://www.legis.state.ak.us/PDF/29/Bills/SB0091Z.PDF. 2 Id.

ACJC Social Impact Bonds Report

nonprofit institutions, however, have finite budgets, which means needs go unmet and new approaches un-tested or under-utilized.

A social impact bond (also referred to as Pay for Success) is an agreement through which several organizations bind together to deliver a program. While not a prerequisite, social impact bonds typically fund preventative efforts that, if successful, obviate government expenses. Structured similar to a performance-based contract, a social impact bond agreement uses private or philanthropic funds to support the delivery of a program with specific outcome measures. Following implementation, an independent evaluation determines whether the outcome measures are met. If the outcome measures are met or exceeded, the government reimburses the private or philanthropic organization, and, depending upon the agreement, may reimburse beyond the initial capital, providing a “return”; if the outcome measures are not met, the private or philanthropic organization, not the government, is liable for the program cost. During this process, an intermediary holds the agreement and helps manage the project. Altogether, there are typically five parties working in concert: the government, the private or philanthropic organization, the service provider, the program evaluator, and the intermediary.

- The government serves as back-end payer, agreeing to reimburse the investors if pre- determined outcomes are achieved.

- The private or philanthropic organization acts as investor, providing capital to finance a program on the condition that the investment is reimbursed when programmatic outcomes are achieved; a rate of return may be triggered if outcomes are exceeded.

- The service provider delivers the intervention, utilizing the funds from the private or philanthropic organization for start-up and program costs.

- The program evaluator verifies the impact of the program while ensuring that the results stem from the intervention, rather than chance or outside factors.

- The intermediary may play various roles depending upon the social impact bond, including project manager, technical support, service provider recruiter, fundraiser, payment facilitator, and contract mediator; all actions and communication between the other partiers are typically routed through the intermediary.

ACJC Social Impact Bonds Report

To accommodate the parties involved, a social impact bond is composed of and operationalized through a series of contracts, including loan/grant contracts, fee-for-service contracts, program evaluation contracts, and reimbursement contracts. These agreements articulate responsibilities, risks assumed and benefits earned upon successful completion of a program. As a result, in addition to being a program funding mechanism, social impact bonds are complex legal agreements, which act as risk transfer vehicles: the expense of an ineffective program, which typically sits with the government or the non-profit service provider, shifts to the investors. As a consequence, the private or philanthropic organization examines the evidence base and determines which, if any, programs merit funding. Where properly aligned and executed, social impact bonds offer preventative programs to populations that might otherwise be neglected until remediation services are necessary.

Examples of Social Impact Bonds in the Field

As a relatively new concept, only a handful of programs using social impact bonds have ended or reached a threshold sufficient for evaluation and reimbursement; four of these follow. Another program designed to reduce recidivism in Massachusetts is included as an example, although the program has not concluded yet. Additionally, although still in the design stage, a program in Alaska is planned that will utilize federal grant funds and a social impact bond to target chronic homelessness and recidivism in Anchorage and the Matanuska-Susitna Borough; an outline of this program is provided.

Peterborough Prison Project3

The Peterborough Prison Project in the United Kingdom was the first to use a social impact bond to finance an intervention. Started in 2010, the goal of the program was to reduce recidivism among offenders who had served sentences of one year or less. The social impact bond provided housing assistance, substance-abuse treatment and mental health services. If the program produced a 10 percent reduction in recidivism among a given cohort during a 2-year follow up period, or a 7.5 percent reduction among all three cohorts, the UK Ministry of Justice would reimburse program investors. In 2014, an independent program evaluation of the first cohort

3 Centre for Social Impact Bonds, Cabinet Office, UK Government. (n.d.). Ministry of Justice: Offenders released from Peterborough Prison. Retrieved from https://data.gov.uk/sib_knowledge_box/ministry-justice- offenders-released-peterborough-prison.

ACJC Social Impact Bonds Report

found an 8.4 percent reduction in recidivism, which was insufficient to trigger repayment on its own. Also in 2014, the UK Ministry of Justice announced that it would discontinue the program two years early, subsuming the recidivism reduction work into a national program.

The New York City Rikers Island Project4

Started in 2012, the New York City Rikers Island Project was designed to reduce the recidivism rate of jailed adolescents through cognitive behavioral therapy. Reimbursement would depend upon the rate of recidivism reduction achieved; triggered initially by an 8.5 percent reduction in recidivism, reimbursement would be capped at a 20 percent reduction. In July 2015, an independent program evaluation found that the program failed to reduce recidivism. As a result, the investors were not reimbursed by the New York City government and the program was discontinued.

Utah High Quality Preschool Project5

Started in 2013, the Utah High Quality Preschool Project was designed to reduce the number of children entering kindergarten who needed special education and decrease the academic achievement gap between low-income and middle/upper-income students. Utah estimated that special education costs the state approximately $2,700 per student per year; the Utah High Quality Preschool Project called for 95 percent of the avoided costs to go to investors. Typical preschool programs reduce the number of students needing special education by 10 or 20 percent, an independent program evaluation found that the Utah program prevented more than 99 percent of the students (109 of 110) from subsequently needing special education.6,7,8 This result

4 City of New York, Office of the Mayor. (August 2, 1012). Fact Sheet: The NYC ABLE Project for Incarcerated Youth - America’s First Social Impact Bond. Retrieved from http://www.nyc.gov/html/om/pdf/2012/sib_fact_sheet.pdf. 5 Goldman Sachs. (November 17, 2015). The Utah High-Quality Preschool Initiative – Pilot and Longitudinal Research Project. Retrieved from http://www.goldmansachs.com/what-we-do/investing-and- lending/impact-investing/case-studies/impact-bond-slc-multimedia/fact-sheet-pdf.pdf. 6 Popper, Nathaniel. (November 3, 2015). Success Metrics Questioned in School Program Funded by Goldman. The New York Times. Retrieved from http://www.nytimes.com/2015/11/04/business/dealbook/did-goldman-make-the-grade.html?_r=0. 7 Savedoff, William and Madan, Janeen. (December 2, 2015). A Social Impact Bond without the Impact? Critics Question Success of Early Childhood Development Program. Center for Global Development. Retrieved from http://www.cgdev.org/blog/social-impact-bond-without-impact-critics-question-success- early-childhood-development-program. 8 Chronicle of Philanthropy. (November 4, 2015). Researchers Question Success of Utah Social-Impact Bond. Retrieved from https://www.philanthropy.com/article/Researchers-Question-Success/234052.

ACJC Social Impact Bonds Report triggered reimbursement of the investor by the state. However, experts have called into question the underlying assumption that all of the students in the program would have needed special education absent the intervention.9,10,11

Massachusetts Chronic Homelessness Pay for Success Initiative12

Started in 2014, the Massachusetts Chronic Homelessness Pay for Success Initiative was designed to reduce chronic individual homelessness. The six year program will provide housing and ancillary services to up to 800 individuals in an effort to manage chronic health issues and reduce emergency service utilization. Success is measured as one year of stable housing; at the close of each program-year, the program evaluator will determine if at least 80% of participants achieved one year of tenancy.13 As of October 2016, the program evaluator determined that the intervention had met its one-year objective and, as a result, the Commonwealth was preparing to repay investors for year-one expenses (PFS stakeholder, email communication, October 10, 2016).

Massachusetts Juvenile Justice Pay for Success Initiative14

In early 2014, Massachusetts launched an initiative to reduce recidivism among juveniles and young adults by expanding an existing program that provides young offenders with life skills and

9 Popper, Nathaniel. (November 3, 2015). Success Metrics Questioned in School Program Funded by Goldman. The New York Times. Retrieved from http://www.nytimes.com/2015/11/04/business/dealbook/did-goldman-make-the-grade.html?_r=0. 10 Savedoff, William and Madan, Janeen. (December 2, 2015). A Social Impact Bond without the Impact? Critics Question Success of Early Childhood Development Program. Center for Global Development. Retrieved from http://www.cgdev.org/blog/social-impact-bond-without-impact-critics-question-success- early-childhood-development-program. 11 Chronicle of Philanthropy. (November 4, 2015). Researchers Question Success of Utah Social-Impact Bond. Retrieved from https://www.philanthropy.com/article/Researchers-Question-Success/234052. 12 Urban Institute. (n.d.) Chronic Homelessness Pay for Success Initiative. Retrieved from http://pfs.urban.org/pfs-project-fact-sheets/content/chronic-homelessness-pay-success-initiative. 13 Nonprofit Finance Fund. (n.d.) Massachusetts Launches the Chronic Individual Homelessness Pay for Success Initiative. Retrieved from http://www.payforsuccess.org/resources/massachusetts-launches- chronic-individual-homelessness-pay-success-initiative. 14 Goldman Sachs. (n.d.) Fact Sheet: The Massachusetts Juvenile Justice Pay for Success Initiative. Retrieved from http://www.goldmansachs.com/our-thinking/trends-in-our-business/massachusetts-social-impact- bond/MA-juvenile-justice-pay-for-success-initiative.pdf.

ACJC Social Impact Bonds Report

job training over a period of four years.15,16 The seven-year initiative is designed to decrease the days the program participants spend in incarceration when compared to similar young offenders who are not in the program. Massachusetts committed up to $27 million and identified a target of at least a 40% decrease in days of incarceration. Payments to investors will be made starting at a rate of at least a 5.2% reduction in incarceration and additional payments will be made for participants’ success in employment. If the project reaches its target, the investors will be repaid their principal plus interest. Payments are scheduled to come out of Massachusetts’ Social Innovation Financing Trust Fund.

Anchorage/Mat-Su Permanent Supportive Housing Demonstration17

In June of 2016, the Anchorage mayor’s office announced that a $1.3 million grant from the U.S. Department of Housing and Urban Development and U.S. Department of Justice will be used to support the implementation of a program to address homelessness financed via a social impact bond. The grant funds will be used to conduct a feasibility analysis and develop the capacity to implement a social impact bond financing structure across the Municipality of Anchorage and Mat-Su Borough (SIB stakeholder, email communication, September 16, 2016). Ultimately funds from the Alaska Mental Health Trust Authority, Mat-Su Health Foundation, and the Rasmuson Foundation will be used to support a Housing First intervention, providing housing and “wrap around comprehensive social services that have demonstrated positive outcomes in assisting chronically homeless people to successfully live in the community.”18 The United Way of Anchorage will act as intermediary.19

15 Field, Anne. (February 7, 2014). Biggest 'Social Impact Bond' In The U.S. Targets Recidivism. Forbes. Retrieved from http://www.forbes.com/sites/annefield/2014/02/07/biggest-social-impact-bond-in-the-u-s- targets-stubborn-recidivism/#5a6651ff25db. 16 Commonwealth of Massachusetts, Executive Department. (January 29, 2014). Massachusetts Launches Landmark Initiative to Reduce Recidivism Among At-Risk Youth. Retrieved from http://www.payforsuccess.org/sites/default/files/massachusetts_press_release_1.29.2014.pdf. 17 Mayor’s Corner, Municipality of Anchorage. (June 28, 2016). United Way of Anchorage Receives Pay of Success Grant to Support Mayor’s Housing and Homeless Action Agenda. Retrieved from http://www.muni.org/Departments/Mayor/PressReleases/Pages/UnitedWayofAnchorageReceivesPayofSu ccessGranttoSupportMayor%E2%80%99sHousingandHomelessActionAgenda.aspx. 18 Id. 19 United Way of Anchorage. (n.d.). Pay for Success - Permanent Supportive Housing Demonstration. Retrieved from http://www.liveunitedanc.org/improving-lives/how-we-work/leading-social-innovation-pay- for-success/.

ACJC Social Impact Bonds Report

Preconditions of Social Impact Bonds

Without rigorous evaluations, much of what is known about social impact bonds must be assembled from practitioners’ experiences designing and implementing these instruments. While there are relatively few examples from which to draw lessons, existing evidence suggests that only a subset of programs are appropriate for social impact bond financing, namely, those that are preventative, reduce budgetary pressure, grow the target population and conform to the evidence base.

Target prevention

In circumstances where the government is acting as a back-end payer, a program financed by a social impact bond typically needs to obviate a government expense. While governments can prioritize a social outcome for reasons other than budgetary savings, in the context of social impact bonds, downstream savings are typically sought as the basis for investor reimbursement. As a consequence, a program financed with a social impact bond should be preventative, where “prevention” is broadly interpreted as any intervention that prevents the need for subsequent government spending.

Reduce budgetary pressure

Programs financed with social impact bonds tend to be more expensive than programs financed directly by government or nonprofit institutions. Research suggests that there are number of reasons for this, including, the need to offer a return to investors based on performance (compensate risk), transaction costs, program evaluation fees, and intermediary fees.20,21 Consequently, except in rare circumstances, social impact bond financing should not seek to supplant directly financed programs, as pressure on a state’s budget would increase.

20 McKay, Kyle. (January 2013). Evaluating Social Impact Bonds as a New Reentry Financing Mechanism: A Case Study on Reentry Programming in Maryland. Annapolis, Maryland: Department of Legislative Services, Office of Policy Analysis. Retrieved from http://mgaleg.maryland.gov/Pubs/BudgetFiscal/2013- Evaluating-Social-Impact-Bonds.pdf 21 United States Government Accountability Office. (September 2015). Pay for Success - Collaboration among Federal Agencies Would Be Helpful as Governments Explore New Financing Mechanisms (GAO-15- 646). Washington, DC: GAO. Retrieved from http://www.gao.gov/assets/680/672363.pdf

ACJC Social Impact Bonds Report

Grow recipient population

While social impact bonds could be used to test novel interventions – rather than replicate and modify existing ones – because investors generally look to mitigate risk through service providers and program models that have a proven track record, these are more difficult to finance.22 Accordingly, practitioners suggest that social impact bonds are best used to supplement existing programs, typically by expanding a program to a larger group or a group not currently impacted.23

Adhere to evidence base

As the expense for an ineffective program sits with the private and philanthropic organization, they, as investors, examine the evidence base to determine which interventions merit funding. While each investor will balance risk versus return differently, programs that most closely conform to the evidence base, have a track record of success, and demonstrate meaningful social impacts will most likely receive funding via a social impact bond.

Avoid creating a dedicated fund

This condition is specific to Alaska. Alaska’s Constitution forbids the creation of a dedicated fund or a contract for state debt. 24 If the Legislature chooses to enter into an obligation using a social impact bond model, it would need to structure the contract so that any state obligation would be subject to future appropriation. This may be less than ideal for some investors. In any event, a social impact bond would likely need to be reviewed by the Department of Revenue’s financial advisor and bond counsel.

22 McKay, Kyle. (January 2013). Evaluating Social Impact Bonds as a New Reentry Financing Mechanism: A Case Study on Reentry Programming in Maryland. Annapolis, Maryland: Department of Legislative Services, Office of Policy Analysis. Retrieved from http://mgaleg.maryland.gov/Pubs/BudgetFiscal/2013- Evaluating-Social-Impact-Bonds.pdf. 23 United States Government Accountability Office. (September 2015). Pay for Success - Collaboration among Federal Agencies Would Be Helpful as Governments Explore New Financing Mechanisms (GAO-15- 646). Washington, DC: GAO. Retrieved from http://www.gao.gov/assets/680/672363.pdf. 24 AK Const. Art. 9, §§ 7, 8.

ACJC Social Impact Bonds Report

Legislative Research Topics

The Alaska Legislature requested the investigation of five topics as a way to explore the potential of using social impact bonds to reduce recidivism rates in Alaska.25 A discussion of these topics follows.

Research Topic 1: Identification and evaluation of grant programs, contracts, and services of the Department of Corrections and the Department of Health and Social Services that may be suitable for social impact financing

Methodology

Alaska Department of Corrections (DOC) and Alaska Department of Health and Social Services (DHSS) interventions were identified through agency websites and documentation. Where possible, programs were matched to the Pew Charitable Trusts’ Results First Clearinghouse Database in order to determine their relative efficacy as established by the evidence base. Matching consisted of comparing program descriptions. Importantly, because DOC and DHSS programs have not been evaluated in the same way that the Results First Clearinghouse has evaluated other programs, these secondary sources must be relied upon and it must be assumed that programs in Alaska and those programs in the clearinghouse are sufficiently similar – in design and implementation. However, actual similarity is unknown and determining it is beyond the scope of this report.

The Clearinghouse used five rating levels: highest rated, second-highest rated, no evidence of effects, mixed effects, and negative effects.26 Interventions that received the highest rating typically were evaluated with a randomized control trial or high-quality quasi-experimental design, and showed a statistically significant positive impact; interventions that received the second-highest rating typically were evaluated with a quasi-experimental design, and showed a positive impact.27 In this evaluation, only those interventions that were rated “highest” or “second-

25 Omnibus Crim Law and Procedure; Corrections, AK SB91, 36 SLA 16 (2016), Retrieved from http://www.legis.state.ak.us/PDF/29/Bills/SB0091Z.PDF. 26 Pew-MacArthur Results First Initiative. (June 23, 2015). Results First Clearinghouse Database – Overview. Retrieved from http://www.pewtrusts.org/en/multimedia/data-visualizations/2015/results-first-clearinghouse- database. 27 Pew-MacArthur Results First Initiative. (June 2015). Results First Clearinghouse Database – User Guide. Retrieved from

ACJC Social Impact Bonds Report highest” were retained. It is important to note that “highest rated” interventions were described in the clearinghouse as either “Effective” or “Strong beneficial” depending on the source; “second-highest rated” interventions were always described as “Promising” (see table below). After screening programs for a match in the Results First Clearinghouse Database and identifying only those programs that demonstrated a positive effect on recidivism, 15 programs remained.

Finally, a benefits to costs ratio was provided for each program by matching interventions to the Washington State Institute of Public Policy benefit-cost database, where available; note that the ratio estimates were developed for Washington State, and, as a result, may not reflect the benefits and costs that exist in Alaska.28 Evaluation of DOC and DHSS programs that are suitable for social impact financing State Evidence Impact on Benefits to Program Agency Based Recidivism Costs Ratio Corrections-Based Adult DOC Promising29 Yes30 $18.3631 Basic/Secondary Education Postsecondary Education - Corrections DOC Promising32 Yes33 $18.3634 Corrections-Based Vocational Training DOC Promising35 Yes36 $12.1337 Parenting Class DOC Effective38 Yes39 N/a

http://www.pewtrusts.org/~/media/assets/2015/06/results_first_clearinghouse_database_user_guide.pdf?la =en. 28 Washington State Institute for Public Policy. (June 2016). Benefit-Cost Technical Documentation - Washington State Institute for Public Policy’s Benefit-Cost Model. Retrieved from http://www.wsipp.wa.gov/TechnicalDocumentation/WsippBenefitCostTechnicalDocumentation.pdf. 29 Crime Solutions. (n.d.). Corrections-Based Adult Basic/Secondary Education. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=21. 30 Id. 31 Washington State Institute for Public Policy. (June 2016). Correctional education (basic or postsecondary) in prison. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/9 32 Crime Solutions. (n.d.). Postsecondary Correctional Education (PSCE). Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=23. 33 Id. 34 Washington State Institute for Public Policy. (June 2016). Correctional education (basic or postsecondary) in prison. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/9 35 Crime Solutions. (n.d.). Corrections-Based Vocational Training Programs. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=24. 36 Id. 37 Washington State Institute for Public Policy. (June 2016). Vocational education in prison. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/6. 38 National Registry of Evidence-based Programs and Practices. (n.d.). Parenting Inside Out. Retrieved from http://legacy.nreppadmin.net/ViewIntervention.aspx?id=345. 39 Id.

ACJC Social Impact Bonds Report

Alaska Reentry Course DOC Promising40 Yes41 N/a Family Violence Intervention Program DOC Effective42 Yes43 ($5.58)44 Sex Offender Treatment - Community DOC Promising45 Yes46 $8.6947 Sex Offender Treatment - Incarcerated DOC Promising48 Yes49 $1.6950 Substance-Abuse Program (Intensive) DOC Strong Yes52 $9.6653 beneficial51 Ignition Interlock Device DOC Effective54 Yes55 N/a Institutional Discharge Project Plus DOC Strong Yes57 $1.5358 beneficial56 Assess, Plan, Identify, Coordinate (APIC) DOC Promising59 Yes60 N/a

40 Crime Solutions. (n.d.). Program Profile: Boston (Massachusetts) Reentry Initiative (BRI). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=42. 41 Id. 42 Crime Solutions. (n.d.). Practice Profile: Interventions for Domestic Violence Offenders: Duluth Model. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=17. 43 Id. 44 Washington State Institute for Public Policy. (June 2016). Domestic violence perpetrator treatment (Duluth-based model). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/86 45 Crime Solutions. (n.d.). Practice Profile: Adult Sex Offender Treatment. Retrieved from https://www.crimesolutions.gov/PracticeDetails.aspx?ID=30. 46 Id. 47 Washington State Institute for Public Policy. (June 2016). Sex offender treatment in the community. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/113. 48 Crime Solutions. (n.d.). Practice Profile: Adult Sex Offender Treatment. Retrieved from https://www.crimesolutions.gov/PracticeDetails.aspx?ID=30. 49 Id. 50 Washington State Institute for Public Policy. (June 2016). Sex offender treatment during incarceration. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/112. 51 What Works in Reentry Clearinghouse. (n.d.). Offender Substance Abuse Pre-Release Program (OSAPP). From https://whatworks.csgjusticecenter.org/program/offender-substance-abuse-pre-release-program- osapp. 52 Id. 53 Washington State Institute for Public Policy. (June 2016). Inpatient/intensive outpatient drug treatment (incarceration). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/189. 54 Crime Solutions. (n.d.). Program Profile: Maryland Ignition Interlock Program. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=63. 55 Id. 56 Work Works in Reentry Clearinghouse. (n.d.). Washington State’s Dangerous Mentally Ill Offender Program. Retrieved from https://whatworks.csgjusticecenter.org/program/washington-states-dangerous- mentally-ill-offender-program. 57 Id. 58 Washington State Institute for Public Policy. (June 2016). Offender Re-entry Community Safety Program (dangerously mentally ill offenders). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/8. 59 Washington State Institute for Public Policy. (June 2016). Offender Re-entry Community Safety Program (dangerously mentally ill offenders). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/8. 59 Crime Solutions. (n.d.). Program Profile: Mental Health Services Continuum Program (Calif.). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=445. 60 Id.

ACJC Social Impact Bonds Report

Alcohol Safety Action Program DHSS Promising61 Yes62 N/a Sobriety 24/7 DHSS Promising63 Yes64 ($0.77)65 Electronic Monitoring DOC Promising66 Yes67 N/a

As with any cost-benefit analysis, not all costs or benefits are included in this analysis, and, as a result, this type of analysis is only one of many factors that may be used to judge the relative value of an intervention. Particularly important when an intervention only affects marginal costs, the higher program costs associated with a social impact bond may create a high hurdle. As an example, in 2013, the Maryland Department of Legislative Services (DLS) evaluated funding reentry services with a social impact bond. DLS calculated that a program that produced a 20 percent reduction in recidivism among 1,250 offenders would be insufficient to close a wing of a prison; given the costs associated with a social impact bond, failure to meet this threshold would mean that a social impact bond would not save the state money.68 These threshold events can be difficult to model. The Results First Initiative, a national project69 currently operational in Alaska,70 is calculating the programmatic benefits and costs of state funded programs. This project will be complete next year, and will be useful in evaluating interventions suitable for social impact bond financing.

Finally, it is important to ensure that an intervention saves the state money – savings that accrue to nonprofit institutions or the federal government create what is known as the “wrong pocket

61 National Registry of Evidence-based Programs and Practices. (November, 2015). Mississippi Alcohol Safety Education Program (MASEP). Retrieved from http://www.nrepp.samhsa.gov/ProgramProfile.aspx?id=21. 62 Id. 63 Crime Solutions. (n.d.). Program Profile: South Dakota’s 24/7 Sobriety Project. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=404. 64 Id. 65 Washington State Institute for Public Policy. (June 2016). Intensive supervision (surveillance only). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/53. 66 Crime Solutions. (n.d.). Program Profile: Electronic Monitoring (Florida). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=230. 67 Id. 68 McKay, Kyle. (January 2013). Evaluating Social Impact Bonds as a New Reentry Financing Mechanism: A Case Study on Reentry Programming in Maryland. Annapolis, Maryland: Department of Legislative Services, Office of Policy Analysis. Retrieved from http://mgaleg.maryland.gov/Pubs/BudgetFiscal/2013- Evaluating-Social-Impact-Bonds.pdf. 69 Information on the Pew-MacArthur Results First Initiative can be found here: http://www.pewtrusts.org/en/projects/pew-macarthur-results-first-initiative. 70 The Alaska Results First project is hosted by The Alaska Justice Information Center at UAA; it’s first report can be found here: https://www.uaa.alaska.edu/academics/college-of-health/departments/justice- center/alaska-justice-information-center/_documents/2016-07-15.results_first_progress_report.pdf.

ACJC Social Impact Bonds Report

problem.” While any program will have spillover effects, some of which are sought, for example, savings captured by members of the community, it is incumbent on practitioners to thoroughly understand and mitigate an intervention’s costs and benefits during the design phase; failure to do so will result in a misalignment of interests.

Research Topic 2: The possibility of private sector investors providing social impact financing

There are private and philanthropic organizations operating in Alaska whose mission overlaps with programs suitable for social impact bond financing. Additionally, there are several organizations which have expressed interest in a program financed by a social impact bond in Alaska; these include, Alaska Mental Health Trust Authority, Mat-Su Health Foundation, and the Rasmuson Foundation. A non-exhaustive list of organizations with a presence in Alaska and a philanthropic interest in recidivism or recidivism-associated interventions follows; the following is in alphabetical order and in no way implies preference.

- Alaska Airlines

o “A limited number of cash grants ranging on average from $5,000 to $15,000 are given to 501(c)(3) non-profit organizations classified as public charities in Alaska, Hawai'i, and Washington. These grants should focus on educational and workforce development efforts that address a unique need or add value to a community” (https://www.alaskaair.com/content/about-us/social- responsibility/alaska-airlines-foundation.aspx).

- Alaska Mental Health Trust Authority

o AMHTA serves individuals with, “mental illness, developmental disability, chronic alcoholism, Alzheimer’s disease or related dementia, [or] traumatic head injury resulting in a permanent brain injury” (http://mhtrust.org/grants/).

o Further, areas of work include, housing and long-term services and supports; workforce development; disability justice; substance abuse prevention and treatment; and, beneficiary employment and engagement (http://mhtrust.org/focus/).

ACJC Social Impact Bonds Report

- Alyeska Pipeline Service Company

o “Alyeska primarily contributes to Alaska-based nonprofits and to Alaska-based arms of national nonprofits that are aligned with the company’s values and philosophy – organizations that support underserved populations, youth, diversity, safety, health and social services, the environment, and workforce development” (http://www.alyeska-pipe.com/AboutUs/DonationGuidelines).

- Arctic Slope Regional Corporation

o “ASRC and its family of companies fund nonprofit and other organizations that provide programs and services in the following focus [area]: . Healthy Community Initiatives” (https://www.asrc.com/Documents/contribution_guidelines.pdf).

- BP

o “BP’s employees contribute to more than 700 community and education organizations and more than 230 youth teams throughout Alaska… To submit a request for funding, please complete the form below. The BP community request form is applicable to eligible Alaska nonprofit organizations only” (http://www.bp.com/en_us/bp-us/where-we-operate/bp-in-alaska/bp-in-the- community.html).

- Bristol Bay Native Corporation

o “In an effort to maximize BBNC’s positive impact in the community, we’ve defined funding priorities that guide our giving across” the following areas: . “Health and Social Services: We fund programs and services that improve the welfare of our shareholders and their families. . “Education, Employment, and Training: BBNC supports events and programs that offer employment, educational and training opportunities for shareholders and their descendants” (http://www.bbnc.net/our- corporation/about/corporate-giving/).

- Chugach Alaska Corporation

o “We seek to sponsor innovative programs, projects and events that: . Address disparities in underserved rural communities within the Chugach region.

ACJC Social Impact Bonds Report

. Improve the well-being of Chugach shareholders and/or the Alaska Native community as a whole, and promote and protect our cultural values. . Promote educational and health initiatives, responsible development, environmental stewardship, safety, economic development, and cultural and ethnic diversity within our communities. . Support federal or state initiatives that align with our organizational objectives. . Support our customers’ missions and/or customer support organizations” (http://www.chugach-ak.com/media/sponsorships).

- ConocoPhilips

o “Historically, an average of approximately 50 percent of our philanthropic dollars goes to programs dedicated to education and youth. Approximately one quarter goes to social services organizations, and the remainder goes to support civic and arts groups as well as environment and industrial safety programs” (http://alaska.conocophillips.com/in- communities/applications/Pages/default.aspx).

- Cook Inlet Region Inc.

o “In reviewing requests, CIRI gives priority to those that specifically benefit Southcentral Alaska… CIRI contributions are focused toward organizations we consider essential for maintaining economic health and quality of life in our communities. We support the following focus areas by building partnerships, promoting volunteerism, and providing resources” (http://www.ciri.com/our- corporation/in-the-community/giving-guidelines/). . “Youth and Education: CIRI supports non-profit organizations that promote and provide educational opportunities and activities for youth (grades K to 12) and gives priority to projects that serve low-income and disadvantaged youth. In Anchorage, CIRI focuses on organizations involved in the effort to increase on-time graduation rates to 90 percent by 2020. . “Homelessness, Hunger and Health: CIRI supports non-profit organizations striving to decrease the number of people that are homeless and hungry

ACJC Social Impact Bonds Report

and provide preventive and primary health care in the areas of heart disease and cancer.”

- First National Bank

o “First National contributes funds to selected non-profit organizations. Each branch manager has the authority to make contributions to local community organizations and events. Additionally, contributions are made at the corporate level by First National's Donations Committee. This Committee meets once each month to review all requests for donations. To allow adequate time for consideration, requests should be received by the bank at least two months before the actual need for funding. First National has developed a set of guidelines to help evaluate requests for support and contributions. Requests will be considered from Alaska based non-profit organizations in four general categories: Community or public service, Health and education, Arts and humanities, Youth and senior citizens” (https://www.fnbalaska.com/about- fnba/charitable-donations).

- Mat-Su Health Foundation

o The Foundation targets the “the health and wellness of Alaskans living in the Mat- Su… In select cases, MSHF may choose to fund health-related projects undertaken by religious, governmental, educational or tribal agencies seeking partners for health-related initiatives that impact the broader Mat-Su community and extend beyond the traditional purpose and functions of these organizations” (http://www.healthymatsu.org/grants/grant-guidelines).

- Northrim BanCorp, Inc.

o “We focus our giving so that we can make meaningful contributions, measure the results of our contributions, and work more closely with the organizations we support. Our focus areas are: Community and economic development, Programs to strengthen low income families” (http://www.northrim.com/home/about/ engagement/community_contributions).

- Rasmuson Foundation

o Program-Related Investments “are financial instruments a foundation can use to support a charitable project or activity. Usually structured as loans, PRIs can also

ACJC Social Impact Bonds Report

be equity investments, linked deposits or loan guarantees. Rasmuson Foundation uses PRIs to achieve charitable outcomes such as increased employment; increased home ownership; neighborhood stability; redevelopment of blighted properties into useful community assets; and increased availability of safe, secure and affordable housing, foods and services. To achieve such outcomes, the Foundation uses PRIs to increase availability of credit and capital to nonprofit intermediaries, banks, community development financial institutions, and small- business developers. The Foundation looks for PRI opportunities in affordable housing, community and economic development, and historic preservation. The Foundation avoids PRI projects that are bankable through traditional venues; could weaken the borrower; that seek debt reduction associated with a current loan; or that support high-risk business ventures not consistent with the Foundation’s core charitable interests” (http://www.rasmuson.org/grants/program-related-investments/).

o “Examples of foundations that have made loan guarantees through program- related investments in recent years include the Rasmuson Foundation, which in 2010 provided credit enhancement through a first loss guarantee for new facility and housing development loans originated in Alaska by nonprofits” (https://nonprofitquarterly.org/2014/10/17/does-pay-for-success-actually-pay-off- the-roi-of-social-impact-bonds/).

- Wells Fargo

o Human Services - “In addition to our significant support of United Way, we consider requests from social and human service organizations for projects with one-time, non-recurring expenses that match our other giving priorities.” These include, community economic development areas, such as, “financial skills education for adults and youth; job development; job placement services; transitional housing services and programs; business management training; revitalization projects for low- and moderate-income areas; economic development initiatives benefiting rural Alaska; [and] affordable housing rehabilitation and construction projects, including multifamily rental housing and single-family homes.” (https://www.wellsfargo.com/about/charitable/ak_guidelines).

ACJC Social Impact Bonds Report

Research Topic 3: Programs operated by nonprofit corporations that could be funded through a social impact financing mechanism

Methodology

Interventions provided by nonprofit agencies in Alaska were identified through previous research from the Criminal Justice Working Group. The agencies in Alaska providing interventions in the identified subject areas include the Alaska Housing Finance Corporation, the Alaska Native Justice Center, and Partners for Progress in Alaska. Where possible, the Alaska programs were matched to the Pew Charitable Trusts’ Results First Clearinghouse Database in order to determine their relative efficacy as established by the evidence base. Matching consisted of comparing program descriptions. It is important to note that, because nonprofit programs generally have not been evaluated, these secondary sources must be relied upon and it must be assumed that programs in Alaska and those programs in the clearinghouse are sufficiently similar – in design and implementation. However, actual similarity is unknown and determining it is beyond the scope of this report.

Pew’s clearinghouse used five rating levels: highest rated, second-highest rated, no evidence of effects, mixed effects, and negative effects. Interventions that received the highest rating typically were evaluated with a randomized control trial or high-quality quasi-experimental design, and showed a statistically significant positive impact; interventions that received the second-highest rating typically were evaluated with a quasi-experimental design, and showed a positive impact. In this evaluation, only those interventions that were rated “highest” or “second- highest” were retained. Important to note that “highest rated” interventions were described in the clearinghouse as either “Effective” or “Strong beneficial” depending on the source; “second- highest rated” interventions were always described as “Promising” (see table below). After screening programs for a match in the Results First Clearinghouse Database and identifying only those programs that demonstrated a positive effect on recidivism, 4 programs remained.

ACJC Social Impact Bonds Report

Evaluation of nonprofit agency programs that are suitable for social impact financing

Evidence Impact on Benefits to Program Based Recidivism Costs Ratio Reentry Services Promising71 Yes72 N/a Housing Assistance Promising73 Yes74 N/a Cognitive Behavioral Therapy Promising75 Yes76 $24.1977 Employment and Job Assistance Strong Yes79 $18.1780 beneficial78

See Research Topic 1 for caveats regarding marginal cost savings and ensuring savings accrue to the state.

Again, programs like those in the clearinghouse are provided in Alaska by agencies such as the Alaska Housing Finance Corporation, the Alaska Native Justice Center, and Partners for Progress in Alaska; insufficient program information is available to formally match these to the programs in the clearinghouse, however. A list of programs provided by these agencies in the areas of reentry services, housing assistance, cognitive behavioral therapy, and employment assistance is available on request.

71 Crime Solutions. (n.d.). Program Profile: Auglaize County (Ohio) Transition (ACT) Program. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=130. 72 Id. 73 What Works in Reentry Clearinghouse. (n.d.). Returning Home – Ohio (RHO) Pilot Program. Retrieved from https://whatworks.csgjusticecenter.org/program/returning-home-ohio-rho-pilot-program. 74 Id. 75 Crime Solutions. (n.d.). Cognitive Behavioral Therapy (CBT) for Moderate- and High-Risk Adult Offenders. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=57. 76 Id. 77 Washington State Institute of Public Policy. (June 2016). Cognitive behavioral treatment (for high and moderate risk offenders). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/10. 78 What Works in Reentry Clearinghouse. (n.d.). EMPLOY (Minnesota). Retrieved from https://whatworks.csgjusticecenter.org/program/employ-minnesota. 79 Id. 80 Washington State Institute of Public Policy. (June 2016). Employment & job training assistance in the community. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/15.

ACJC Social Impact Bonds Report

Research Topic 4: Independent evaluators that could determine whether performance targets for a nonprofit corporation funded by social impact financing are met at the end of an agreed-upon time frame

The following is a non-exhaustive list of Alaska-based firms or organizations that cite “program evaluation” or a variant as a service; the following is in alphabetical order and in no way implies preference.

- Organization: Agnew::Beck

o Type: Consulting, Private o Location: Anchorage and Boise, ID o Services and Skill Set: Program Evaluation and Assessment, “Demonstrating the effectiveness of a program or project is critical to ensuring that the chosen strategies are achieving the desired results. To help secure future funding and public support, evaluation explains and justifies the value of a program. Agnew::Beck designs program evaluations to meet each organization’s needs, funder requirements and the specified budget. We are committed to engaging a wide range of stakeholders and bringing in multiple perspectives through the design and implementation process. This ensures that findings are representative and comprehensive and that the organization can use them to improve their programs in a meaningful way” (http://agnewbeck.com/capabilities/services/program-evaluation-assessment/).

o Areas of Expertise: Community Facility Development; Economic Development; Energy; Environmental Analysis & Planning; Housing / Affordable Housing; Land Use & Urban Design; Public, Behavioral & Community Health; Rural Development; Tourism, Recreation, & Open Space Planning; Workforce Development

o Previous Work: Yupiit Piciryarait Cultural Center and Museum Native Archive Project Evaluation, http://agnewbeck.com/2012/11/avcp-museum-program- evaluation-final-report/.

- Organization: EvaluLogic

o Type: Consulting, Private o Location: Anchorage o Services and Skill Set: Program Evaluation: “An effective evaluation should identify: A logical evaluation strategy; Key programmatic components; Key

ACJC Social Impact Bonds Report

evaluation questions; Key targets for evaluation activities; Misalignment of program activities and intended outcomes; Measureable metrics; Strategies for measurement and analysis; Program improvement options” (http://www.evalulogic.com/pages/what-we-do.php#program-evaluation).

o Previous Work: Alaska INBRE-3 Evaluation Plan - Alaska INBRE-3 is a continuation of previous NIH funding to stimulate biomedical research at the University of Alaska through infrastructure, faculty development, and student training. This 5- year, $15 million program spans 3 campuses and multiple departments; AK BioPREP - laska BioPREP was a National Institutes of Health/SEPA-funded program designed to incorporate biotechnology activities into both rural and urban high school classrooms across the state; http://www.evalulogic.com/pages/our- work.php.

- Organization: Goldstream Group

o Type: Consulting, Private o Location: Fairbanks o Services and Skill Set: Program Evaluation: “The Goldstream Group works with clients to develop evaluations to improve programs, demonstrate program effectiveness, and inform decision-making. Our staff is skilled in both quantitative and qualitative methods of data collection and analysis, and we use multiple methods to provide programs with formative and summative data” (http://www.goldstreamgroup.com/home-2/program-evaluation).

o Areas of Expertise: Specializes “in the development and evaluation of educational, health, and social service programs” (http://www.goldstreamgroup.com/what-we-do).

o Previous Work: Expanding Our Horizons – “Expanding Our Horizons, funded by the U.S. Department of Education to develop English language skills of students in the Yukon-Koyukuk School District. To document the impact of the project on teachers and students, Goldstream Group designed teacher satisfaction surveys, interviewed teachers; conducted observations of project activities, and analyzed student standards based assessment data. Goldstream Group provided annual written reports and completed U.S. Department of Education reporting forms” (http://www.goldstreamgroup.com/expanding-our-horizons); Next Generation Evaluation – “In the fall of 2008, the Juneau School District began to implement its Next Generation: Our Kids, Our Community high school education plan. The plan

ACJC Social Impact Bonds Report

includes several components, including theme-based academies, 9th grade smaller learning communities, and expanded extracurricular activities. Goldstream Group’s five-year evaluation includes written surveys with teachers and students, focus groups with students, individual interviews with teachers, and statistical analysis of student cohort data. Goldstream Group provides a written annual report and presentation to the Board of Education” (http://www.goldstreamgroup.com/next-generation-evaluation-2).

- Organization: McDowell Group

o Type: Consulting, Private o Location: Anchorage and Juneau o Services and Skill Set: Program Evaluation and Needs Assessments – “Our experienced staff uses surveys, confidential interviews, best-practices research, and statistical and economic analysis to measure program needs, capacity and impacts. We work closely with clients to understand and address the implications of our findings” (http://www.mcdowellgroup.net/our-services/program- evaluation-needs-assessments/).

o Areas of Expertise: “Social and public issues such as health, education and the delivery of social services are a significant component of McDowell Group’s professional practice. Our firm has completed more than 200 projects concerning varied subjects such as education service delivery, education cost allocations, rural education, alcohol abuse, the guardianship system, senior services, and critical access evaluation for hospitals. To help our clients enhance the impact of their programs, we have employed a variety of research and consulting methods including compilation and analysis of economic and socio- economic data, community and client surveys, confidential interviews with industry and public leaders, facilitation of public meetings, and strategic planning” (http://www.mcdowellgroup.net/areas-of-expertise/health-education- social-services/).

o Previous Work: Alaska Youth Courts Evaluation and Impact Assessment, http://www.globalyouthjustice.org/uploads/Alaska_Youth_Courts_Evaluation.pdf; Alaska 21st Century Learning Centers Statewide Evaluation, https://education.alaska.gov/21cclc/pdf/statewide_eval_report.pdf.

ACJC Social Impact Bonds Report

Other entities that could provide independent evaluation services include: Innovative Consulting Inc., out of Fairbanks, and research units within the University of Alaska system. Commission staff will provide more information on these possibilities upon request.

Research Topic 5: Whether federal funding is available for independent evaluators participating in social impact funding

As of July 2016, there were no federal funds available (currently or in the pipeline) to support the evaluation component of social impact bonds.

However, Harvard Kennedy School Government Performance Lab provides several types of assistance to state and local partners who have agreed to embark on a social impact bond project. Currently, applications are being accepted for pro-bono assistance with social impact bond project development; selection criteria include the following: 1. The potential to advance the PFS [social impact bond] field by applying the model in new ways, in new geographic areas, or in new policy fields. 2. A high level of commitment from government leaders. 3. The potential impact of the project. 4. Readiness to engage in project activities. Applications for assistance are due by due March 15, 2017 (http://govlab.hks.harvard.edu/apply).

Outside of this application period, the Harvard Kennedy School Government Performance Lab provides 12-15 months of pro-bono technical services to “guide government partners through the programmatic, budgetary, regulatory and procurement processes of project development”; applications for this technical assistance are reviewed on an on-going basis (http://govlab.hks.harvard.edu/apply).

Conclusion and Recommendations

Given the relatively untested nature of social impact bonds, it is difficult to say whether the state of Alaska should embrace them. Social impact bonds shift the risk of financing preventative programs to private entities, requiring relatively little initial expenditure by the state. There would be some necessary initial costs involved in contracting and administration, and care must be

ACJC Social Impact Bonds Report

taken to draft social impact bond agreements to avoid the pitfalls and adhere to the best practices listed above.

Social impact bonds also require a future commitment by the state to reimburse the private entities for their investment. One potential problem inherent in that future commitment is whether future administrations will have desire and ability to fund the reimbursements; essentially the state must commit to including these reimbursements in budgets that have not yet been written. Even if the state is able to set funds aside, the money set aside for possible future reimbursement might need to come from other projects and programs that have already been established.

The state must also be able to ensure that the savings from the contracted programs will go to the state, rather than other entities. If not, the state may end up committing to reimbursements that will not be paid for in future savings. The Alaska Results First Initiative, which is currently evaluating Alaska’s recidivism reduction programs, will be able to provide a better understanding of the cost- benefit impact of these programs in the near future.

The Commission therefore recommends to the legislature that it not actively solicit social impact bonds, as that would require expending state resources on a relatively unknown mechanism. If, however, a private entity were to submit a proposal to the state for a social impact bond, the Commission recommends accepting such a proposal, provided that the proposed contract is thoroughly vetted to ensure that it will result in savings to the state and will not drain funds from other needed projects.

ACJC Social Impact Bonds Report

Works Cited

Alaska Department of Corrections. (January 8, 2016). Alaska Department of Corrections - Programs and Services. Retrieved from http://www.correct.state.ak.us/doc/programs- services.pdf

Centers for Disease Control and Prevention. (2015). Motor Vehicle Prioritizing Interventions and Cost Calculator for States (MV PICCS) 2.0. Retrieved from http://www.cdc.gov/motorvehiclesafety/calculator/index.html

Centre for Social Impact Bonds, Cabinet Office, UK Government. (n.d.). Ministry of Justice: Offenders released from Peterborough Prison. Retrieved from https://data.gov.uk/sib_knowledge_box/ministry-justice-offenders-released-peterborough-prison

Chronicle of Philanthropy. (November 4, 2015). Researchers Question Success of Utah Social- Impact Bond. Retrieved from https://www.philanthropy.com/article/Researchers-Question- Success/234052

City of New York, Office of the Mayor. (August 2, 1012). Fact Sheet: The NYC ABLE Project for Incarcerated Youth - America’s First Social Impact Bond. Retrieved from http://www.nyc.gov/html/om/pdf/2012/sib_fact_sheet.pdf

Commonwealth of Massachusetts, Executive Department. (January 29, 2014). Massachusetts Launches Landmark Initiative to Reduce Recidivism Among At-Risk Youth. Retrieved from http://www.payforsuccess.org/sites/default/files/massachusetts_press_release_1.29.2014.pdf

Crime Solutions. (n.d.). Cognitive Behavioral Therapy (CBT) for Moderate- and High-Risk Adult Offenders. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=57

Crime Solutions. (n.d.). Corrections-Based Adult Basic/Secondary Education. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=21

Crime Solutions. (n.d.). Corrections-Based Vocational Training Programs. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=24

Crime Solutions. (n.d.). Program Profile: Electronic Monitoring (Florida). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=230

Crime Solutions. (n.d.). Postsecondary Correctional Education (PSCE). Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=23

Crime Solutions. (n.d.). Practice Profile: Adult Sex Offender Treatment. Retrieved from https://www.crimesolutions.gov/PracticeDetails.aspx?ID=30

Crime Solutions. (n.d.). Practice Profile: Interventions for Domestic Violence Offenders: Duluth Model. Retrieved from http://www.crimesolutions.gov/PracticeDetails.aspx?ID=17

Crime Solutions. (n.d.). Program Profile: Auglaize County (Ohio) Transition (ACT) Program. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=130

Crime Solutions. (n.d.). Program Profile: Boston (Massachusetts) Reentry Initiative (BRI). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=42

ACJC Social Impact Bonds Report

Crime Solutions. (n.d.). Program Profile: Maryland Ignition Interlock Program. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=63

Crime Solutions. (n.d.). Program Profile: Mental Health Services Continuum Program (Calif.). Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=445

Crime Solutions. (n.d.). Program Profile: South Dakota’s 24/7 Sobriety Project. Retrieved from http://www.crimesolutions.gov/ProgramDetails.aspx?ID=404

Field, Anne. (February 7, 2014) Biggest 'Social Impact Bond' In The U.S. Targets Recidivism. Forbes. Retrieved from http://www.forbes.com/sites/annefield/2014/02/07/biggest-social-impact-bond- in-the-u-s-targets-stubborn-recidivism/#5a6651ff25db

Goldman Sachs. (n.d.) Fact Sheet: The Massachusetts Juvenile Justice Pay for Success Initiative. Retrieved from http://www.goldmansachs.com/our-thinking/trends-in-our- business/massachusetts-social-impact-bond/MA-juvenile-justice-pay-for-success-initiative.pdf

Goldman Sachs. (November 17, 2015). The Utah High-Quality Preschool Initiative – Pilot and Longitudinal Research Project. Retrieved from http://www.goldmansachs.com/what-we- do/investing-and-lending/impact-investing/case-studies/impact-bond-slc-multimedia/fact- sheet-pdf.pdf

Mayor’s Corner, Municipality of Anchorage. (June 28, 2016). United Way of Anchorage Receives Pay of Success Grant to Support Mayor’s Housing and Homeless Action Agenda. Retrieved from http://www.muni.org/Departments/Mayor/PressReleases/Pages/UnitedWayofAnchorageReceiv esPayofSuccessGranttoSupportMayor%E2%80%99sHousingandHomelessActionAgenda.aspx

McKay, Kyle. (January 2013). Evaluating Social Impact Bonds as a New Reentry Financing Mechanism: A Case Study on Reentry Programming in Maryland. Annapolis, Maryland: Department of Legislative Services, Office of Policy Analysis. Retrieved from http://mgaleg.maryland.gov/Pubs/BudgetFiscal/2013-Evaluating-Social-Impact-Bonds.pdf

National Registry of Evidence-based Programs and Practices. (November, 2015). Mississippi Alcohol Safety Education Program (MASEP). Retrieved from http://www.nrepp.samhsa.gov/ProgramProfile.aspx?id=21

National Registry of Evidence-based Programs and Practices. (n.d.). Parenting Inside Out. Retrieved from http://legacy.nreppadmin.net/ViewIntervention.aspx?id=345

Nonprofit Finance Fund. (n.d.) Massachusetts Launches the Chronic Individual Homelessness Pay for Success Initiative. Retrieved from http://www.payforsuccess.org/resources/massachusetts- launches-chronic-individual-homelessness-pay-success-initiative

Omnibus Crim Law and Procedure; Corrections, AK SB91, 36 SLA 16 (2016), Retrieved from http://www.legis.state.ak.us/PDF/29/Bills/SB0091Z.PDF

Pew-MacArthur Results First Initiative. (June 23, 2015). Results First Clearinghouse Database – Overview. Retrieved from http://www.pewtrusts.org/en/multimedia/data- visualizations/2015/results-first-clearinghouse-database

ACJC Social Impact Bonds Report

Pew-MacArthur Results First Initiative. (June 2015). Results First Clearinghouse Database – User Guide. Retrieved from http://www.pewtrusts.org/~/media/assets/2015/06/results_first_clearinghouse_database_user_gu ide.pdf?la=en

Popper, Nathaniel. (November 3, 2015). Success Metrics Questioned in School Program Funded by Goldman. The New York Times. Retrieved from http://www.nytimes.com/2015/11/04/business/dealbook/did-goldman-make-the- grade.html?_r=0

Savedoff, William and Madan, Janeen. (December 2, 2015). A Social Impact Bond without the Impact? Critics Question Success of Early Childhood Development Program. Center for Global Development. Retrieved from http://www.cgdev.org/blog/social-impact-bond-without-impact- critics-question-success-early-childhood-development-program

United States Government Accountability Office. (September 2015). Pay for Success - Collaboration among Federal Agencies Would Be Helpful as Governments Explore New Financing Mechanisms (GAO-15-646). Washington, DC: GAO. Retrieved from http://www.gao.gov/assets/680/672363.pdf

United Way of Anchorage. (n.d.). Pay for Success - Permanent Supportive Housing Demonstration. Retrieved from http://www.liveunitedanc.org/improving-lives/how-we- work/leading-social-innovation-pay-for-success/

Urban Institute. (n.d.) Chronic Homelessness Pay for Success Initiative. Retrieved from http://pfs.urban.org/pfs-project-fact-sheets/content/chronic-homelessness-pay-success-initiative

Washington State Institute of Public Policy. (June 2016). Cognitive behavioral treatment (for high and moderate risk offenders). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/10

Washington State Institute for Public Policy. (June 2016). Correctional education (basic or postsecondary) in prison. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/9

Washington State Institute for Public Policy. (June 2016). Domestic violence perpetrator treatment (Duluth-based model). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/86

Washington State Institute of Public Policy. (June 2016). Employment & job training assistance in the community. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/15

Washington State Institute for Public Policy. (June 2016). Intensive supervision (surveillance only). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/53

Washington State Institute for Public Policy. (June 2016). Offender Re-entry Community Safety Program (dangerously mentally ill offenders). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/8

Washington State Institute for Public Policy. (June 2016). Outpatient/non-intensive drug treatment (incarceration). Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/190

Washington State Institute for Public Policy. (June 2016). Sex offender treatment during incarceration. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/112

ACJC Social Impact Bonds Report

Washington State Institute for Public Policy. (June 2016). Sex offender treatment in the community. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/113

Washington State Institute for Public Policy. (June 2016). Benefit-Cost Technical Documentation - Washington State Institute for Public Policy’s Benefit-Cost Model. Retrieved from http://www.wsipp.wa.gov/TechnicalDocumentation/WsippBenefitCostTechnicalDocumentation .pdf

Washington State Institute for Public Policy. (June 2016). Vocational education in prison. Retrieved from http://www.wsipp.wa.gov/BenefitCost/Program/6

What Works in Reentry Clearinghouse. (n.d.). EMPLOY (Minnesota). Retrieved from https://whatworks.csgjusticecenter.org/program/employ-minnesota

What Works in Reentry Clearinghouse. (n.d.). Offender Substance Abuse Pre-Release Program (OSAPP). Retrieved from https://whatworks.csgjusticecenter.org/program/offender-substance- abuse-pre-release-program-osapp

What Works in Reentry Clearinghouse. (n.d.). Returning Home – Ohio (RHO) Pilot Program. Retrieved from https://whatworks.csgjusticecenter.org/program/returning-home-ohio-rho-pilot- program

Work Works in Reentry Clearinghouse. (n.d.). Washington State’s Dangerous Mentally Ill Offender Program. Retrieved from https://whatworks.csgjusticecenter.org/program/washington-states- dangerous-mentally-ill-offender-program